On 5.6.2012 17:15, Les Mikesell wrote: > On Tue, Jun 5, 2012 at 9:47 AM, Gene Poole <gene.poole-swIuAdtPLFgAvxtiuMwx3w at public.gmane.org> wrote: >> >> 3. During the installation I instituted software raid-1 along with >> LVM. It all worked perfectly until I got to the Boot Loader screen and >> instead of using the default (/dev/sda) I used what I thought would work >> in a raid-1 environment IF one of the hard drives went bad. I chose to >> install the boot loader in /dev/md0. Of course it would not boot. I >> recovered from this by going through the install and choosing to upgrade a >> existing installation and it allowed me to place the boot loader in >> /dev/sda. What that leaves me with is a raid-1 environment that works >> great as long as /dev/sda remains. How do I fix that??? > > The MBR isn't mirrored, so you just have to install grub on the other > drive, usually by executing grub There is an open bugzilla https://bugzilla.redhat.com/show_bug.cgi?id=799501 I am irritated regularly about this. Because It WORKSFORME. Of course the mbr isn't mirrored but at installation time it is written to both disks. I wipe the mbr of sdb, reinstall and after that $ dd if=/dev/sdb count=1 bs=512|strings shows something like GRUB, on *both disks* in /root/anaconda-ks.cfg I find bootloader --location=mbr --driveorder=sda,sdb ... in the original kickstart there is bootloader --location=mbr ... I did it per kickstart, is this a problem with manual installs only? Or only with upgrade mode (Gene said he used upgrade mode)? But then again there is this bugzilla... -- Kind Regards, Markus Falb -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 307 bytes Desc: OpenPGP digital signature URL: <http://lists.centos.org/pipermail/centos/attachments/20120606/dde1c196/attachment-0005.sig>